School Children Theories on ChatGPT: Educational Perspectives of the Implementation of Artificial Intelligence in Social Robots

This chapter explores the increasing significance of child-social robot collaboration in educational contexts, emphasizing the integration of large language models (LLMs) such as ChatGPT into social robots to enhance children’s trust and interaction. Rooted in a long tradition of research inspired by Piaget and Turkle’s studies on children’s interactions with social media, the investigation aims to understand children’s perceptions of AI, emotions, and trust. An experiment was conducted involving children interacting with ChatGPT to promote reasoning about AI. The procedure included: introducing AI concepts, brainstorming and writing questions, interacting with ChatGPT, analyzing responses, and reflecting on the chatbot’s reasoning. The activities were designed to stimulate critical thinking about AI’s capabilities and limitations. This approach seeks to help children better understand AI and develop critical thinking skills regarding systems like ChatGPT.